case Study
Tamratyare
POLICYANALYSISBYWILLIAMDUNN1.pptChapter 1
The Process of Policy Analysis
Compare and contrast methods of policy analysis and evaluation
Understand the role of methods in creating and transforming policy-relevant information
Explain how methods are related to phases of the policy-making process described in the last module
Recognize the importance of documentation and communication
Understand the role of external factors in conducting policy analyses and program evaluations
Discuss opportunities to conduct policy papers based on the use of policy-analytic methods
Chapter 1
Learning Objectives
Chapter 1
Chapter 1
Chapter 1
Factors influencing the practice of policy analysis and evaluation:
Organizational cultures
Problem-solving styles
Institutional incentives
Organizational structures
Time constraints
Organizational learning
Chapter 1
Factors
Passive vs. active political culture/intuitive vs. thinking-sensing vs. feeling/positive and negative sanctions for challenging status quo/
Chapter 1
- Agenda Setting. Stakeholders in and outside government compete to put problems on the government agenda.
- Formulation. Potential solutions are formulated by staff in ministries, legislatures, executive offices.
- Adoption. A policy is officially adopted by an executive, legislative, or judicial organ.
- Implementation. The policy is carried out by administrative agencies within ministries and departments.
- Assessment. The outcomes of policies are monitored and evaluated by special agencies.
- Adaptation. Policies are changed to fit previously unknown circumstances.
- Succession. Policies are continued with new goals.
- Termination. Policies and institutions are terminated.
Phases of Policy Making
Agenda setting
Policy formulation
Policy adoption
Policy implementation
Policy assessment
Policy adaptation
Policy succession
Policy termination
Problem structuring
Forecasting
Recommending
Monitoring
Evaluation
Problem resolving
Problem unsolving
Problem solving
Chapter 1
Policy Analysis in the Policymaking Process
Chapter 2
Policy Analysis in the Policymaking Process
When we think about policies that are developed to respond to policy issues, we should be able to explain the process by which these policies are made and implemented.
How do problems get the attention of policy makers?
How are policies made and implemented?
How are they monitored and evaluated?
How are policies maintained, changed, or terminated?
How can methods of policy analysis help improve this process?
Chapter 2
The Problem
The process of public policy making is a political process based on the exercise of political power and legal authority
Power and authority are exercised by executive, legislative, and judicial bodies at local, national, and international levels
The process of policy making has multiple phases ordered in time—chains, cycles, detours, short-circuits
Policies are made more or less quickly, by a few or many persons, with small or (rarely) large changes
Four models (I-IV) help us understand the process of making and implementing policies
The process of policy analysis helps improve the process of policy making by providing policy-relevant information that is useful in the policy process.
Chapter 2
Key Points
Policy agenda setting
Policy formulation
Policy adoption
Policy implementation
Policy evaluation
Policy adaptation
Policy succession
Policy termination
Chapter 2
Phases of Policy Making
Agenda setting – cigarette labeling – EU, EC, EP, ECJ require transposition and harmonization of regulations and directives of EU
Formulation – Policy challenging 2003 Census developed as a new law—voted down-- Ban on standardized tests to enter universities proposed and never reaches Congress
Adoption – Parliament amends (rebalances) budget to include funds for transportation
Implementation – Implementation of Law on Local Government implemented by Vice President and by Ministry of Local Self Government – new employees hired, training programs identified, tenders let
Evaluation – News media reports on corruption, so does Transparency International, so does General Secretariat
Adaptation – Directives of the acquis shaped to fit local conditions – same with “equalization fund” for local governments
Succession – NHTSA keeps 55 mph limit after it saves lives
Termination – Civil servants agency abolished in Slovakia; Unite for Reform abandoned in Serbia; OTA abandoned in US
Model I: Rational Actor
Model II: Organizational Process
Model III: Bureaucratic Politics
Model IV: Interrupted Equilibrium
Chapter 2
Four Models of Policy Making
Analogy —The policy process is like an economic enterprise or company In which a CEO chooses an investment alternative that earns the greatest net profit. Example: Sn = (1 + r)n
Rule —The greater the benefits of an alternative, and the less the costs, the more likely the alternative will be chosen.
Chapter 2
Model I: Rational Actor
Policy makers agree on a problem
They identify objectives
They list all policy alternatives
They predict all outcomes
They determine utility/value of outcomes
They choose the optimal alternative
Chapter 2
Characteristics of Model I
Analogy— The policy process is like an unending debate in which participants adjust their positions because they are forced to negotiate and compromise.
Rule — Partisan policy makers mutually adjust their policies, so that policies at one time, t, are only marginally different from policies at a later time, t+1.
Chapter 2
Model II: Organizational Process
Policy makers adjust objectives after tradeoffs
Policies made at the margins of status quo
Policies based on a limited set of alternatives
Changes in policies occur in small increments
Problems reconstructed when new information becomes available
Analysis and evaluation occur throughout society in a process that is fragmented and disjointed
Policies involve small steps to remedy a problem rather than cure them completely with radical steps
Chapter 2
Characteristics of Model II
Analogy —The policy process is like a battle among inhabitants of relatively isolated islands, each of which has its own program and its own ways of rewarding and punishing its own islanders.
Rule — “Where you stand depends on where you sit.” The favored policy of a bureaucratic leader depends on the agency or ministry in which she sits.
Chapter 2
Model III: Bureaucratic Politics
Analogy —Policy making is like biological evolution. Most policies involve small, relatively small changes over long periods of time. There is a stable, dynamic equilibrium among competing policies—but from time to time there are abrupt and perhaps irreversible changes.
Rule —Periodically, external shocks produce new political beliefs and attitudes, including fear, and these result in large and abrupt changes in policies.
Chapter 2
Model IV: Interrupted Equilibrium
Chapter 2
Q & A
Use widely respected methods of policy analysis to provide more and better information in each phase of policy making
Translate this information and analysis into a language that is understandable to others
Prepare written policy documents including memos, regulatory impact assessments (RIAs), policy issue papers, and research reports on potential solutions to problems
Use oral briefings, meetings, conversations, and conferences to communicate the contents of policy documents
Chapter 2
Improving the Policy Making Process
Chapter 2
Chapter 2
Improving the Policy Making Process
RETROSPECTIVE: What happened and was it worthwhile?
PROSPECTIVE: What will happen and will it be worthwhile?
PROBATIVE: What problem should be solved?
DEMONSTRATIVE: What is the solution to the problem?
Chapter 2
Questions Answered by Methods
Chapter 2
Impact Matrix (Scorecard)
Chapter 2
Spreadsheet
Chapter 2
Influence Diagram
Chapter 2
Analysis of a Policy Argument
Policy agenda setting Structuring policy problem
Policy formulation Forecasting policy outcomes
Policy adoption Recommending preferred policy
Policy implementation Monitoring policy outcomes
Policy evaluation Evaluating policy performance
Policy adaptation Recommending adapted policy
Policy succession (Re)commending existing policy
Policy termination Recommending no policy
Chapter 2
Policy Analysis In The Policy Process
Agenda setting – cigarette labeling – EU, EC, EP, ECJ require transposition and harmonization of regulations and directives of EU
Formulation – Policy challenging 2003 Census developed as a new law—voted down-- Ban on standardized tests to enter universities proposed and never reaches Congress
Adoption – Parliament amends (rebalances) budget to include funds for transportation
Implementation – Implementation of Law on Local Government implemented by Vice President and by Ministry of Local Self Government – new employees hired, training programs identified, tenders let
Evaluation – News media reports on corruption, so does Transparency International, so does General Secretariat
Adaptation – Directives of the acquis shaped to fit local conditions – same with “equalization fund” for local governments
Succession – NHTSA keeps 55 mph limit after it saves lives
Termination – Civil servants agency abolished in Slovakia; Unite for Reform abandoned in Serbia; OTA abandoned in US
POLICY ANALYST
POLICYMAKING PROCESS
POLICY DOCUMENTS
POLICY COMMUNICATIONS
Dissemination
Utilization
Analysis
POLICY INFORMATION
Documentation
Chapter 2
Cognitive styles
Analytic roles
Institutional incentive systems
Time constraints
Professional socialization
Multidisciplinary teamwork
Organizational cultures
Political constraints
Chapter 2
Factors Influencing the
Practice of Policy Analysis
The practice of policy analysis refers to the actual processes of reasoning used by analysts (logics-in-use). These should be contrasted with formal representations of reasoning that are to some extent methodological idealizations (logical reconstructions).
Chapter 2
Three Dimensions of Utilization
Composition
(users)
Effects
(type)
Scope
(Information)
Chapter 3
Structuring Policy Problems
Understand the process of problem structuring
Contrast relatively well-structured, moderately structured, and ill-structured problems
Describe Type III errors in policy analysis
Learn how to conduct a stakeholder analysis
Use different methods of problem structuring with a problem of your choice
Learning Outcomes
*
*
We fail more often because we define the wrong problem, than because we get the wrong solution. We commit Type III errors: Defining the wrong problem.
Type III errors can be fatal—”Wrong problem, wrong solution!”
Problems are formed by the interaction of thought and external environments—they are interdependent, subjective, artificial, and dynamic.
Problems are wholes not merely parts—an analysis of parts of a problem may miss the whole.
Policy makers tend to avoid rather than benefit from conflicting perspectives—they prefer consensus.
Complex Problems
*
Mouse analogy
Poverty is not an original sin
Pyramid analogy
Disciplinary blinders
Three Types of Problems
Chapter 2
| Problem Element | Well Structured | Moderately Structured | Ill Structured |
| STAKEHOLDERS | One | Several | Many |
| ALTERNATIVES | Known | Partially Known | Mostly Unknown |
| OUTCOMES OF ALTERNATIVES | Known | Partially known | Mostly Unknown |
| PROBABILITIES OF OUTCOMES | Objective & Determined | Objective & Uncertain | Subjective & Risky |
| VALUE (UTILITY) OF OUTCOMES | Unanimity | Consensus | Conflict |
*
Simon’s Creative Architect: Custom house without standard plans
Modern Diogenes of Sinop: Looking for causes, outcomes, impacts, and objectives
Ceteris paribus—Drunk and key under lamppost
Indian proverb: It is darkest under the lamp (EIII)
Many of the most important probability judgments are subjective/personal
A stakeholder is any individual or group that affects/and or is affected by a policy. Stakeholders may be identified by name and title, sampled with little error, prioritized, and queried indirectly or by simulation for their perspectives of a problem.
Internal versus external
Formal position
Reputation for influence
Functional role
International versus domestic
Identifying Stakeholders
*
Chapter 2
*
Personal Perspective. Individual interests, values, character …
Institutional Perspective. Bureaucratic politics, incrementalism, interrupted equilibrium …
Technical Perspective. Benefit-cost analysis, econometrics, microeconomic policy analysis…
Chapter 2
Problem Structuring
with Multiple Perspective Analysis
*
Chapter 4
Forecasting Expected Policy Outcomes
“The future matters to everyone because that is where we will all be spending the rest of our lives.” Nicholas Rescher, Predicting the Future: An Introduction to the Theory of Forecasting (1998)
Prediction is crucially important to public policy because it is our sole window on the future—and that is where the success and failure of policies will be known.
Key Ideas
*
Contrast projections, predictions, and conjectures
Understand how institutional contexts affect forecast accuracy
Compare and contrast goals and objectives of forecasts
Distinguish extrapolative and theoretical forecasting
Make point and interval forecasts
Analyze a case in environmental justice where political conflicts affect forecast accuracy
Learning Objectives
*
| GOAL | OBJECTIVE |
| General purpose: “increase citizen participation” | Specific purpose: “increase participation at meetings by 20%” |
| Formal definition: ”quality health care means accessibility to treatment” | Operational definition: “quality care refers to doctors per 1000 persons” |
| Time period not specified: “in the period ahead” | Time period specified: “in the 2004 fiscal year” |
| Primarily qualitative: “adequate number of licensed physicians” | Primarily quantitative: “an additional 400 licensed physicians” |
*
Extrapolation
Prediction
Expert judgment
Forms of Forecasts
*
*
The Logic of Extrapolation
*
The Logic of Prediction
*
The Logic of Expert Judgment
*
Recommending Preferred Policies
Chapter 5
Learning Objectives
- Distinguish policy recommendation from other methods of policy analysis
- Describe six criteria used to choose policies
- Contrast comprehensive rationality and disjointed incrementalism
- Describe different types of policy rationality
- List and illustrate steps in conducting benefit-cost and cost-effectiveness analyses
- Apply benefit-cost analysis to a case study of U.S. and European efforts to save lives gasoline by setting maximum speed limits
*
*
Criteria Used to Choose Policies
Adequacy
Efficiency
Effectiveness
Equity
Responsiveness
Appropriateness
*
Comprehensive Rationality—A Naive Model of Policy and Management
Agree on a problem
Identify and rank objectives
List all policy alternatives
Forecast outcomes
Determine utility of outcomes
Choose the optimal alternative
*
Partisan Mutual Adjustment—A
More Realistic Model
Make policies (policies are made) at the margin of the status quo
Policy makers (consider) a limited set of alternatives
Policy makers (seek) incremental changes
(They) limit the number of outcomes considered for each alternative
(They) limit the number of outcomes considered for each alternative
*
(They) adjust objectives to policies after partisan tradeoffs
(They) reconstruct problems when new information becomes available
(They) repeat analysis and evaluation in a series of sequential chains
(They) use analysis and evaluation to remedy existing ills, not to cure problems based on preconceived goals
(They) recognize that analysis and evaluation occur throughout society in a process that is fragmented or disjointed
*
Types of Policy Rationality
Economic rationality —efficiency of 2+ alternatives
Technical rationality —effectiveness in achieving outcomes
Legal —conformity/compliance to rules
Social —institutionalization of rights
Substantive —wise or prudent choices among different forms of rationality
Erotetic —discovery of rationality is part of process of being rational
*
Conducting a Benefit-Cost Analysis
Identify alternatives
Specify objectives
Identify target groups and beneficiaries
List all benefits and costs
Collect data for analysis
Discount benefits and costs to present value
Select criterion of choice
Compare benefits and costs
Make recommendation
*
Discounting Benefits and Costs
Discount rate: The rate at which money can be borrowed, or the rate at which money invested elsewhere will accumulate. A rate of 10 percent (0.10) is the average discount rate over a number of years.
Discount factor: The factor by which a future sum of money is discounted back to its present value. The discount factor is the reciprocal of the rate of interest—1/1+r .
*
Present Value of Benefit Stream of $100 Calculated at 10 Percent Discount Rate
| Year | Future Value (fv) | Discount Rate (r) | Number Periods (n) | Discount Factor (df) | Present Value (pv) |
| 2003 | $110.00 | 0.10 | 1 | 1/(1+0.10)1 = 0.909 | $110.00 |
| 2004 | $121.00 | 0.10 | 2 | 1/(1+0.10)2 = 0.826 | $100.00 |
| 2005 | $133.10 | 0.10 | 3 | 1/(1+1.0)3 = 0.751 | $100.00 |
*
Benefits and Costs of the 55 mph
Speed Limit
COSTS
Hours Driving
H = [1.04VM1973/S1974 – VM1973/S1973] x R = 1.95 billion
H = [VM1973/S1974 – VM1973/S1973] x R
= 1.72 billion
Value of Hours
$5.05/hr (average wage) = $9.85 billion
$1.67/hr (survey) = $2.89 billion
*
Costs of Enforcement
$.8 million
$12 million
BENEFITS
Gasoline Saved
$0.718 cents (price support) = $2,500 billion
$0.528 cents (market price) = $1,442 million
*
Lives saved
$1,297.7 million
$998 million
Injuries
$942.3 million
$722 million
Property damage
$472 million
$236 million
A Net Benefits = $2,321.2 B/C = 1.8
B Net Benefits = - $6,462 B/C = .345
*
Monitoring Observed
Policy Outcomes
Chapter 6
Social systems Lorenz curve
accounting Gini Index
Regression discontinuity Purchasing power
Social experimentation Random innovation
Social auditing Quasi-experimentation
Research and practice Evaluability assessment
synthesis Internal validity
Threats to validity External validity
Interrupted time-series Control-series
Current Euros Constant Euros
Key Terms and Concepts
*
Learning Objectives
Distinguish monitoring from other methods
List the main functions of monitoring
Contrast outcomes and impacts
Distinguish approaches to monitoring
List threats to internal and external validity
Perform interrupted time-series and control-series analysis with SPSS
Compare the U.S. and European speed limit cases
Participate in an in-class Delphi analysis
Chapter 6
*
*
The Importance of Time
*
Why Monitoring Is Important
It is not that we have so many well-designed policies. Rather, we have more well-designed policies than we have ways to monitor them. Without monitoring, we cannot know a good policy from a bad one—or whether the policy is a policy at all.
*
An Unmonitored Policy May
Conceal A Disabled Vehicle
Failing to monitor the outcomes of a policy is like counting the amount of gasoline a car has consumed without seeing how far it has traveled.
*
Four Functions of Monitoring
Compliance—Are laws on local government consistent with EU requirements?
Auditing—Are revenues intended for local communities reaching them?
Accounting—Are policies on educational reform producing qualified students?
Explanation—Are outcomes of a policy caused by the policy, or by other factors?
*
Approaches to Monitoring
Social Systems Accounting
Social Auditing
Research and Practice Synthesis
Policy Experimentation
*
Social Systems Accounting
Housing—Area per person (square meters)
Average Life Expectancy
Quality Adjusted Life Years
Income Distribution (Gini Index)
Air Pollution Index (parts per million)
Lead Concentration Index (blood concentration)
Persons in Mental Hospitals
Persons Below Poverty Line
*
Social Auditing With User Surveys
Policy-Program Specification—What goals, objectives, and resources constitute the policy?
Collection of Available Information—What information is available on inputs, processes, outputs and impacts?
Policy Modeling—What causal mechanisms link inputs and processes to outputs and impacts?
Evaluability Assessment—Is the policy clear enough and unambiguous to know what to monitor?
Collection of New Information—What new information needs to be collected?
*
Research and Practice Synthesis
Synthesis of research on planned change, communication of innovations, social marketing strategies (journals and books)
Synthesis of published and unpublished policy documents (memos, reports, statistics)
Synthesis of reported cases of change, innovation, and reform (case survey analysis)
*
Policy Experimentation
Randomized Policy Experiments—Like randomized clinical trials in medicine, randomized policy experiments involve the direct manipulation of an intervention and random selection of participants and random assignment of participants to an intervention and control group.
Natural Policy Experiments (also called “quasi-experiments”—Random selection and assignment are not possible or ethical, but there are intervention and control groups.
*
Threats to Validity (Rival Hypotheses) When Conducting Policy Experiments
Statistical Conclusion Validity
Internal Validity
External Validity
Construct Validity
Context Validity
*
Statistical Conclusion Validity
The approximate validity of inferences about covariation, in any of its statistical forms, between an intervention and one or more of its presumed outcomes. The approximate statistical conclusion validity of claims based on the classical linear model of regression analysis is diminished to the extent that assumptions of linearity, homoscedasticity, uncorrelated errors, and other statistical requirements are violated.
*
The approximate validity of inferences about the existence of a causal relation between an intervention (the presumed cause) and one or more outcomes (the presumed effects), however statistically valid. The approximate internal validity of an inference relating cause and effect will be diminished to the extent that statistical covariation is weak or absent, the temporal precedence of the presumed cause is ambiguous or unknown, and other plausible causes are not eliminated.
Threats to Internal Validity
*
The approximate validity of inferences about the generalizability of internally valid causal relations to other contexts, settings, persons, groups, interventions, and outcomes. The approximate external validity of a generalized causal inference will be diminished to the extent that the effects of an intervention in one context or setting are undetectable in other contexts or settings, the original intervention is sufficiently complex (or diffuse) that its replication elsewhere is in doubt, and the outcomes are weak or absent among other persons or groups.
Threats to External Validity
*
The approximate validity of inferences about abstract categories, concepts, or labels used to characterize properties of contexts, settings, persons, groups, interventions, or outcomes, and one or more of their relations. The approximate construct validity of such categories, concepts, or labels will be diminished for reasons that include inadequate formal and operational definitions of constructs, failure to examine relations among multiple overlapping constructs, and failure to recognize and account for the effects of procedures for measuring and observing constructs on the existence of the constructs.
Threats to Construct Validity
*
The approximate validity of inferences about the representativeness (ecological typicality) of causally relevant constructs, and hypotheses formed by these constructs, in specific social, spatial, and temporal contexts. The approximate context validity of constructs and hypotheses will be diminished to the extent that they are unrepresentative of the conceptual ecology of persons who affect or are affected by an intervention.
Threats to Context Validity
*
Interrupted Time-Series
Policy Intervention
*
3.bin
Some Major Threats to Validity
History
Maturation
Instability
Instrumentation
Testing
Mortality
Selection
Regression toward the mean
Violated assumptions of statistical tests
*
Evaluating Policy Performance
Chapter 7
Values Are Central to Policy Analysis
Learning Objectives
Compare and contrast monitoring and evaluation
Describe and illustrate criteria for evaluating policy performance
Contrast causal evaluation, official evaluation, and participative evaluation
Describe how ethics affect market-centered and polis-centered perspectives of policy and management
Explain the process of reasoning about values
Show how valuation affects the evaluation of fiscal decentralization policy in Macedonia
Criteria for Evaluating Policy Performance
Effectiveness
Efficiency
Adequacy
Equity
Responsiveness
Appropriateness
Three Approaches to Evaluation
| Approach | Aims | Assumptions | Example |
| Causal Evaluation | Analysts determine outcomes | Values can be described but not justified | Field experiment |
| Official Evaluation | Policymakers determine objectives | Values can be stated and need no justification | Summativeevaluation |
| Participative Evaluation | Stakeholders determine objectives | Values can be stated and need no justification | Evaluability assessment |
Two Perspectives of Values
MARKET-CENTERED
Individual as focus
Self-interest primary motivation
Performance through private competition
Society governed by fixed and impersonal economic laws (“laws of matter”)
Personal decision criteria are individual interest maximization and cost minimization
Change occurs through material exchange and the satisfaction of aggregate individual interests
Public administration is unproductive (“bureaucracy”)
POLIS-CENTERED
Community as focus
Public and self-interest are primary motivations
Performance through cooperation and publicly managed competition
Society governed by laws that are subject to human change (“laws of passion”)
Personal decision criteria are loyalty, public commitment, and individual interest
Information relatively incomplete and subjective
Change occurs through persuasion, alliances, and the satisfaction of public and community interests
Public administration can be productive (“public trust”)
Reasoning About Values
Rule 1 Some municipalities should receive resources from former municipalities that have been consolidated or eliminated in the reform, and no longer have membership rights.
Rule 2 More efficiently managed municipalities should get a larger share because they use funds more productively, as indicated by their market size, per capita income, or some other basis for ranking.
Rule 3 Municipalities that have suffered past discrimination should get a larger share.
Rule 4 Municipalities that have not received their share because of political interference by central government should get a larger share.
Rule 5 Municipalities should be allowed to refuse revenues as a means to gain freedom from central control.
Rule 6 Municipalities should retain revenues generated through savings and investment.
Rule 7 Revenues should be distributed by a lottery, where every municipality has an equal chance of being selected for revenues.
Rule 8 A group of municipalities, or technical experts selected by the municipalities, should vote on the distribution of revenues.
Group Simulation
Break into three groups. Assume that your group
is an expert commission responsible for making a
recommendation about the formula that should be used
for the local government “equalization fund” in Macedonia.
The groups should use these rules:
- Group I: Rules 1, 2, and/or 3
- Group II: Rules 4 and/or 5
- Group III: Rules 6, 7, and/or 8
Use the structural model of argument to develop a well-justified
recommendation.
Value duality Evaluability assessment
Effectiveness User survey analysis
Efficiency Values
Equity Norms
Responsiveness Teleological (utilitarian)
Appropriateness Deontological
Evaluation Valuation
Normative ethics Metaethics
Multiattribute utility Terminal values
analysis Instrumental values
Key Terms and Concepts
Developing Policy Arguments
Chapter 8
Learning Objectives
- Understand the origins of argumentation analysis as an approach to policy
- Describe elements of the structural model of argument
- Contrast types of policy claims
- Explain the dynamics of policy argumentation
- Distinguish different modes of policy argumentation
- Identify formal and informal fallacies of reasoning
- Apply methods of argumentation analysis to a case of intervention in the Balkans
Graduate Center for Public Policy and Management
Background
Historical origins in Aristotle’s Rhetoric and Thucydides’ Melian Dialogues
Modern development in Stephen Toulmin’s The Place of Reason in Ethics (1948) and The Uses of Argument (1958)
Toulmin’s structural model of argument and his theory of practical reasoning are highly influential
The “argumentative turn” in policy studies represents a shift from formal to practical reasoning, and a movement from the idea of “proof” to that of “justification”
Graduate Center for Public Policy and Management
Argumentation analysis has been used to expose the misuse of language in political ideologies and in the social and behavioral sciences.
Policy analysts in universities and in corporations and government departments have been influenced by the structural model of argument.
The use of argumentation analysis is a reaction to “logical positivism” and notions that quantification is an ideal language, pure objectivity is an attainable goal, and science is value free.
The main purpose of argumentation analysis is to fight dogma, facilitate open, critical discourse, and protect democratic institutions now threatened by the “scientization” of policy.
Graduate Center for Public Policy and Management
The Six Elements of the Structural Model of Argument
[I]nformation: Is the information relevant to the issue and does it provide grounds for the claim?
[C]laim: What conclusion or recommendation can we reach on the basis of the information?
[Q]ualifier: How plausible or true is the claim?
[W]arrant: What assumptions or arguments justify moving from information to claim?
[B]acking: What additional assumptions or arguments establish the truth or plausibility of the warrant?
[R]ebuttal: Are there special circumstances or conditions that weaken Q by challenging the plausibility of W, B, or I?
Graduate Center for Public Policy and Management
Structure and Dynamics of Argumentation
Graduate Center for Public Policy and Management
Types of Policy Claims
Designative (“The end of the Cold War was due to President Reagan’s ‘get tough’ policy with the Soviet Union).”
Evaluative (“The distribution of income has become more and more inequitable. This is unjust”)
Advocative (“We recommend that the Department of Health and Human Services oversee the implementation of universal health care.”)
Graduate Center for Public Policy and Management
Modes of Policy Argument
Authority
Method
Generalization
Classification
Cause
Sign
Motivation
Intuition
Analogy-metaphor
Parallel case
Ethics
Graduate Center for Public Policy and Management
Argumentation from Authority
Reasoning is based on warrants having to
do with the achieved or ascribed statuses
of producers of knowledge. For example:
experts, insiders, scientists, consultants,
gurus, power brokers. Footnotes and references are authoritative arguments.
(“The National Academy of Sciences
concluded that the temperature of the earth
will increase by 1 degree F. every 11 years.”)
Graduate Center for Public Policy and Management
Argumentation from Authority
Graduate Center for Public Policy and Management
Argumentation from Method
Reasoning is based on warrants about the status of methods used to produce knowledge. The focus is on the status or “power” or “robustness” of methods or their results, rather than authoritative persons. Examples include statistical, econometric, qualitative, and ethnographic methods.
Graduate Center for Public Policy and Management
Argumentation from Method
Graduate Center for Public Policy and Management
Argumentation from Generalization
Reasoning is based on similarities between samples and populations, or on qualitative comparisons. The assumption is that what is true of members of a sample will also be true of members of the population not included in the sample. Example: Random samples of n 30 are taken to be representative of the unobserved (and often unobservable) population from which the sample is drawn.
Graduate Center for Public Policy and Management
Argumentation from Generalization
Graduate Center for Public Policy and Management
Argumentation from Classification
Reasoning has to do with membership in a defined class. The reasoning is that what is true of the class of persons or events described in the warrant is also members of the class. Example: The ideological argument that because a country has a socialist economy, it must be undemocratic, because all socialist systems are undemocratic.
Graduate Center for Public Policy and Management
Argumentation from Classification
Graduate Center for Public Policy and Management
Argumentation from Cause
Reasoning is about generative powers ("causes") and their consequences ("effects"). Claims are based on social or economic laws stating or implying invariant relations between causes and effects, or on observations that a policy always has a certain effect. Most argumentation in the social and natural sciences is based on reasoning from cause. Example: “Privatization improves governmental efficiency.”
Graduate Center for Public Policy and Management
Argumentation from Cause
Graduate Center for Public Policy and Management
Argumentation from Sign
Reasoning is based on signs, or indicators. The presence of a sign indicates the presence of an event, because the sign and what it refers to occur together. Examples: Indicators of institutional performance such as “organizational report cards,” “best practices,” “benchmarks,” or indicators of economic performance such as “leading economic indicators”—they are sometimes used as causes. But indicators are not causes, because causality must satisfy requirements not expected of signs.
Graduate Center for Public Policy and Management
Argumentation from Sign
Graduate Center for Public Policy and Management
Argumentation from Motivation
Reasoning is based on the motivating power of goals, values, or intentions in shaping behavior. Example: A claim that citizens will support the strict enforcement of pollution standards is based on reasoning that, since citizens are motivated by the desire to achieve the goal of clean air and water, they will act to offer their support.
Graduate Center for Public Policy and Management
Argumentation from Motivation
Graduate Center for Public Policy and Management
Argumentation from Intuition
Reasoning is based on the conscious or preconscious cognitive, emotional, or spiritual states of producers of knowledge. Example: The awareness that an advisor has some special insight, feeling, or "tacit knowledge" may serve as a reason to accept his judgment.
Graduate Center for Public Policy and Management
Argumentation from Analogy-Metaphor
Reasoning is based on similarities between the relations found in a given case and the relations described in a metaphor or analogy. Example: The claim that a government should “quarantine” a country by interdicting illegal drugs—with the illegal drugs seen as an “infectious disease”—is based on reasoning that, since quarantine has been effective in cases of infectious diseases, interdiction will be effective in the case of illegal drugs. “Garbage cans,” “primeval policy soups.”
Graduate Center for Public Policy and Management
Argumentation from Analogy-Metaphor
Graduate Center for Public Policy and Management
Argumentation from Parallel Case
Reasoning is based on similarities among two or more policies. Example: A local government should adopt a particular tax code, because a parallel policy was successfully implemented under similar conditions in another country.
Graduate Center for Public Policy and Management
Argumentation from Parallel Case
Graduate Center for Public Policy and Management
Argumentation from Ethics
Reasoning is based on the rightness or wrongness, goodness or badness, of policies or their consequences. Claims may be based on moral principles of a “just” or “good” society, or on ethical norms prohibiting lying in public life. Many arguments about economic benefits and costs involve unstated or implicit ethical reasoning. Example: “A just social state is one in which one person is better off and no one is worse off; or the winners can compensate the losers, at least in principle.”
Graduate Center for Public Policy and Management
Argumentation from Ethics
Graduate Center for Public Policy and Management
Communicating Policy Analysis
Chapter 9
Learning Objectives
Understand the role of documentation and communication in promoting the use of policy analysis
Describe elements of an oral briefing or presentation
Identify principles for communicating ideas to different groups and individuals
Use criteria for effective communication to evaluate oral briefings
Use Presentation Planner to organize and present a briefing that communicates results of an analysis of lead poisoning
*
Policy-Relevant Information is Produced By Methods of Analysis
*
Information is Utilized Through Processes of Documentation and Communication
POLICY ANALYST
POLICY PROCESS
POLICY DOCUMENTS
POLICY BRIEFINGS
Communication
Utilization
Analysis
POLICY INFORMATION
Documentation
*
Basic and Applied Policy Analysis
| Characteristic | Basic Analysis | Applied Analysis |
| ORIGIN OF PROBLEMS | Academics | Practitioners |
| COMMUNICATIONS | Journal article | Memo or issue paper |
| NATURE OF DATA | Primary data | Secondary data |
| AIM OF ANALYSIS | Improve theory | Improve practice |
| LOCUS OF INCENTIVES | Universities | Governments |
*
Criteria for Assessing Policy Memos and Other Documents
Economy of style
Clarity
Directness
Understandability
Organization
Attention-Getting
Costs to reader
*
Tasks in Policy Documentation
Synthesis
Evaluation
Organization
Translation
Simplification
Visualization
Display
Summary
*
Steps in Writing a Policy Memo
State question(s) the memo will answer
Review prior attempts to solve problem
Diagnose scope, severity, and causes of problem
Identify goals and objectives
Compare alternatives according to benefits, costs, and constraints
State conclusions and/or recommendations
Provide attachments (as appropriate)
*
Elements of An Issue Paper
Letter of Transmittal
Executive Summary
Background of Problem
Scope, Severity, Causes of Problem
Description of Alternatives—Goals and Objectives
Analysis of Alternatives—Costs and Benefits
Conclusions and/or Recommendations
References/Sources
Appendices
*
Elements of a Policy Briefing
Opening and Problem Statement
Background and Objectives
Findings Related to Objectives
Methods of Research and Analysis
Data Supporting the Findings
Recommended Solutions to Problem
Questions from Audience
Closing and Summary
*
Criteria for Evaluating Policy Briefings and Presentations
Effectiveness of elements of briefing
Appropriateness of briefing to characteristics of audience
Logic, organization, and flow
Use of slides or other visual displays
Ability to capture attention
Benefits and costs to audience
*
How to use Presentation Planner to communicate the results of a policy analysis titled:
“When Statistics Count: Revising
the EPA Lead Standard,” by David
L. Weimer and Aidan L. Vining
*
WHAT
Options for Revising the EPA Lead Standard
WHY
Demonstrate the use of Presentation Planner
BY WHOM
William N. Dunn and Colleagues
TO WHOM
Participants in GSPIA 2009
WHERE
Graduate Center for Public Policy and Management
WHEN
November 11, 2003, 1800-2100h
*
Good evening colleagues. It is good to see you again. It is also an honor and privilege to have Dr. Joseph Josifoski with us tonight. Please stand, Joe.
Friends, with your help this evening, we hope to improve the analysis we have been conducting for the past 10 months. Our analysis examines options for regulating emissions of atmospheric lead, which as you know has become a severe public health problem.
Opening (1 of 5)
*
Opening (2 of 5)
Before we begin I would like to introduce the members of our research group: Ana Zabevska, Ph.D., Bekim Imeri, M.D., Meri Kostovska, J.D., and Andrija Aleksoski, M.D.
Members of our group represent four areas of expertise: survey research and sampling, biostatistics, econometrics, and epidemiology.
My name is Bill Dunn and I direct the Environmental Protection Agency.
*
Opening (3 of 5)
Here is the agenda for this evening:
- Background and main objectives of the analysis (Dr. Zabevska)
- Findings with respect to each objective (Dr. Imeri)
- Methods and data supporting findings
(Dr. Kostovska) - Recommendations for action
(Dr. Aleksoski)
*
Opening (4 of 5)
We ask that you hold your questions until after we have finished our presentations. I will serve as moderator, and presenters will respond to questions in their own areas of expertise.
Unless there are questions, let me now turn to Dr. Zabevska.
*
Notepad for Opening
1. Remember to acknowledge Dr. Josifoski and his support. Call him “Joe” to indicate he is a friend, and to stress that the meeting is informal. Have him stand. You may want to initiate applause.
2. Limit questions to clarifications of the agenda. Ask the audience to hold substantive questions for the end of the presentations.
*
Supplement 1: Measuring Outcomes and Impacts
*
Policy Intervention
Patterns of Causality in Time Series
*
6.bin
Conditions Required to
Make Causal Inferences
Condition X precedes condition Y in time X O
Condition X is correlated with condition Y rx.y > 0 (+/-)
Conditions other than X do not affect condition Y
r x.y = rx.y|z
*
Research Designs Help Make
Plausible Causal Inferences
X
I
Yb
X
Ya
~ X
Yb
Ya
II
III
R
R
R
R
R
R
IV
V
VI
VII
Yt
YT
X
~ X
Ya
Ya
X
~ Xi
YT,C
Ya
*
Strengths of Quasi-Experimental Designs
Activity theory of causation
Recognition of systemic complexity
Financial, political, and ethical feasibility
Rarity of true experiments
Availability of resources (www.economagic.com, www.fedstats.gov, www.census.gov, www.eurostat, Statisticki Godisnjak/ Bilten)
*
Extended Time Series
I O1 O2 O3 O4 O5 O6 O7
*
Interrupted Time Series
I O1 O2 O3 X O4 O5 O6 O7
*
Control Series
I O1 O2 O3 X O4 O5 O6 O7
II O1 O2 O3 ~X O4 O5 O6 O7
*
Problems with Interrupted Time Series
Incremental diffusion of programs with no sharp cutting points
Multiple programs operating at same time
Lack of detailed knowledge of program activities
Insufficient observations in time series
Unknown time intervals due to delays in implementing programs
Multiple rival explanations of outcomes
*
Interrupted Time-Series Analysis
Helps Detect Causality
Policy Intervention
*
15.bin
Some Outcome Indicators
Housing—Area per person (square meters)
Average Life Expectancy
Quality Adjusted Life Years
Persons Below Poverty Line
Income Distribution (Gini Index)
Air Pollution Index (parts per million)
Lead Concentration Index (blood concentration)
Persons in Mental Hospitals
Average Test Scores
Sales or Market Share
Votes Cast for Candidates
Foreign Direct Investment in MKD
Number of newly licensed foreign companies
*
The Odds Ratio Measures Effect Size
Example--It is believed that more highly educated voters tend to vote for Democratic candidates in the U.S. Here is a sample of voters who voted in the 1992 Presidential Election. How would a policy of producing more Masters and Ph.D.graduates affect the outcome of elections?
Clinton
Bush and Perot
Less than Masters
Masters or Ph.D. Degree
797 (0.48)
82 (0.42)
857 (0.52)
111 (0.58)
P
1-Q
1-P
Q
1,654 (1.0)
193 (1.0)
P / 1-P = 0.92
Q / 1-Q = 1.38
ODDS RATIO = 1.38 / 0.92 = 1.5
*
The Standardized Mean Difference
Measures Effect Size
Example—Between 1987 ands 1989 the maximum speed limit in 40 of the 50 states of the U.S. was increased from 55mph to 65mph. The paired t-test, which involves a change in means from t0 to t+1 (Note: Observations in any time series are not independent), was used to test the null hypothesis that there is no statistically significant difference (p = 0.05) between traffic fatalities before (1987) and after (1989) the speed limit was raised to 65 mph in 40 states. The speed limit was kept at 55 mph for 10 states. What does the following test show about the effects of removing the old (55mph) policy?
texp = mean fatality rate after the policy - mean fatality rate before the policy / pooled standard deviation
= -0.23 / 0.35 = -0.66
tcon = mean fatality rate after the policy – mean fatality rate before the policy / pooled standard deviation
= -0.07 / 0.10 = -0.70
*
Guidelines for Interpreting
Standardized Effect Sizes
- 0.80-0.99 strong
- 0.60-0.79 moderate to strong
- 0.40-0.59 moderate
- 0.20-0.39 weak to moderate
- 0.00-0.19 negligible to weak
NOTE: The practical significance of an effect size depends on the social costs of being wrong.
*
Other Measures of Effect Size
Identical Units of Measure. Benefits and costs in constant value of a currency, unemployment rates, percent of budget variance, performance appraisal scale.
Established Norms. Dietary intake of vitamins compared with minimum (RDA) required daily amount, international test scores, percent above poverty line, percent below a “living wage.”
Average Effect Sizes. Average correlations in political science and sociology range from r = 0.20 to r = 0.30. Average internal consistency reliabilities for mental health inventories, placement examinations, and other instruments involving high risk of being wrong are r > 0.95.
Coefficient of Variation. The standard deviation divided by the mean
(CV = s / m) . This is the percent variability divided by the mean. A standard deviation, s, of 16 with a mean of 100 is the same as a standard deviation, s, of 96 with a mean of 600. The variability of large municipal budgets can be compared with smaller ones.
*
Pooled t-Test. The outcome mean after the intervention subtracted from the outcome mean before the intervention, divided by the pooled standard deviation. NOTE: The observations before and after the intervention are not independent and therefore the pooed t-test must be used.
x2 – x1 / sqrt [Sp (1/n1) + (1/n2)]
Standard (z) Scores. An individual score subtracted from the mean of the distribution divided by the standard deviation.
z = x - m / s . An individual score of 116 from a distribution with a mean of 100 and a standard deviation of 16 is the same as an individual score of 348 from a distribution with a mean of 300 and a standard deviation of 48. Individual scores measured with two different scales, or from two different distributions, can be directly compared.
*
Interrupted Time-Series
With Two Observations
*
16.bin
Interrupted Time-Series
With Three Observations
*
17.bin
Extended Time-Series
With Interruption
*
18.bin
Extended Time-Series
With Interruption
*
19.wmf
YEAR
2000
1998
1996
1994
1992
1990
1988
1986
1984
1982
1980
1978
1976
1974
1972
1970
1968
1966
FATALITIES
60000
50000
40000
30000
Control Series With Interruption
*
20.bin
Changes in Fatalities Per Mile
Correlated with Economic Factors
*
21.bin
Control Series With Interruption:
Fatality Rates in Europe and the US
*
22.bin
Annual Changes in Fatality Rate
and Miles Driven, 1913-2000
*
23.bin
Group Problem
Examine the extended time-series graphs showing the observed fatality rate, the predicted fatality rate, and European Commission target for 2010.
1. Explain how interrupted time-series analysis might result in a different predicted fatality rate. Is the observed fatality rate a valid predictor of fatalities in future?
2. Explain how control-series analysis might change the Commission’s 2010 target fatality rate. Is the target realistic?
*
Forecast—EU Fatality Rate by 2010
*
ETSC’s proposal for a challenging but realistic EU-wide target was based on examination of long term trends in traffic and casualties and current levels of activity to provide a forecast to 2010.
Annual fatality totals and traffic volumes for all Member States were brought together between 1970-1997. A model of the trends was applied and on the basis of that, a forecast was made of the number of fatalities in 2010.
Chart1
| 1980 | 1980 | 1997 |
| 1981 | 1981 | 1998 |
| 1982 | 1982 | 1999 |
| 1983 | 1983 | 2000 |
| 1984 | 1984 | 2001 |
| 1985 | 1985 | 2002 |
| 1986 | 1986 | 2003 |
| 1987 | 1987 | 2004 |
| 1988 | 1988 | 2005 |
| 1989 | 1989 | 2006 |
| 1990 | 1990 | 2007 |
| 1991 | 1991 | 2008 |
| 1992 | 1992 | 2009 |
| 1993 | 1993 | 2010 |
| 1994 | 1994 | |
| 1995 | 1995 | |
| 1996 | 1996 | |
| 1997 | 1997 |
data
| 1. The data | Data summed over the 15 EU Member States | ||||||||||||||||||
| The graphs show how the number of fatalities in the 15 EU Member States has fallen | traffic | fatalities | |||||||||||||||||
| over a period of rapid traffic growth. In order to forecast how the number of fatalities | (bn veh-km) | ||||||||||||||||||
| may change in future, we need to explain past changes, then predict what would | 1970 | 1085 | 77989 | ||||||||||||||||
| happen if similar changes were to continue in future. | 1971 | 1170 | 79335 | ||||||||||||||||
| 1972 | 1245 | 81768 | |||||||||||||||||
| To see how this can be done, click on the next sheet rates | 1973 | 1309 | 77705 | ||||||||||||||||
| 1974 | 1299 | 70279 | |||||||||||||||||
| 1975 | 1360 | 69510 | |||||||||||||||||
| 1976 | 1418 | 69829 | |||||||||||||||||
| 1977 | 1476 | 68374 | |||||||||||||||||
| 1978 | 1554 | 67508 | |||||||||||||||||
| 1979 | 1602 | 65057 | |||||||||||||||||
| 1980 | 1645 | 64063 | |||||||||||||||||
| 1981 | 1661 | 61062 | |||||||||||||||||
| 1982 | 1705 | 59750 | |||||||||||||||||
| 1983 | 1730 | 59451 | |||||||||||||||||
| 1984 | 1797 | 56693 | |||||||||||||||||
| 1985 | 1840 | 52695 | |||||||||||||||||
| 1986 | 1936 | 54780 | |||||||||||||||||
| 1987 | 2044 | 52750 | |||||||||||||||||
| 1988 | 2154 | 55091 | |||||||||||||||||
| 1989 | 2255 | 56023 | |||||||||||||||||
| 1990 | 2351 | 56426 | |||||||||||||||||
| 1991 | 2425 | 56013 | |||||||||||||||||
| 1992 | 2485 | 52784 | |||||||||||||||||
| 1993 | 2528 | 48265 | |||||||||||||||||
| 1994 | 2575 | 46537 | |||||||||||||||||
| 1995 | 2627 | 46111 | |||||||||||||||||
| 1996 | 2680 | 43521 | |||||||||||||||||
| 1997 | 2742 | 43413 | |||||||||||||||||
| 1998 | 42699 | 41% | |||||||||||||||||
| 1994-98 average | 44456 | 44% | |||||||||||||||||
| 25000 | |||||||||||||||||||
| 1997 | 4245 | 1149 | |||||||||||||||||
| 2010 | 4939 | 1362 | |||||||||||||||||
| 116% | 119% | ||||||||||||||||||
| 27150 | 40725 | ||||||||||||||||||
| 0.6782797228 |
data
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
rates
| 2. The fatality rate | ||||||||||||||||
| It is natural to expect the number of fatalities to be roughly proportional to the volume of | ||||||||||||||||
| traffic. The graph shows the EU fatality rate - the number of fatalities per billion vehicle-km | fatality rate: | fatality number: | ||||||||||||||
| travelled. The rate has fallen very regularly for many years, as shown by the red line. | actual | modelled | actual | modelled | ||||||||||||
| 1970 | 71.87 | 72.09 | 77989 | 78226 | ||||||||||||
| 1971 | 67.80 | 67.62 | 79335 | 79131 | ||||||||||||
| 1972 | 65.66 | 63.43 | 81768 | 78991 | ||||||||||||
| 1973 | 59.34 | 59.50 | 77705 | 77911 | ||||||||||||
| 1974 | 54.08 | 55.82 | 70279 | 72529 | ||||||||||||
| 1975 | 51.10 | 52.36 | 69510 | 71225 | ||||||||||||
| 1976 | 49.25 | 49.11 | 69829 | 69627 | ||||||||||||
| 1977 | 46.33 | 46.07 | 68374 | 67994 | ||||||||||||
| 1978 | 43.44 | 43.21 | 67508 | 67162 | ||||||||||||
| 1979 | 40.61 | 41.00 | 65057 | 65672 | ||||||||||||
| 1980 | 38.95 | 38.89 | 64063 | 63964 | ||||||||||||
| 1981 | 36.77 | 36.89 | 61062 | 61261 | ||||||||||||
| 1982 | 35.04 | 35.00 | 59750 | 59680 | ||||||||||||
| 1983 | 34.36 | 33.20 | 59451 | 57438 | ||||||||||||
| 1984 | 31.55 | 31.50 | 56693 | 56597 | ||||||||||||
| 1985 | 28.64 | 29.88 | 52695 | 54975 | ||||||||||||
| 1986 | 28.30 | 28.34 | 54780 | 54867 | ||||||||||||
| 1987 | 25.81 | 26.89 | 52750 | 54956 | ||||||||||||
| 1988 | 25.58 | 25.51 | 55091 | 54940 | ||||||||||||
| 1989 | 24.85 | 24.20 | 56023 | 54554 | ||||||||||||
| 1990 | 24.00 | 22.95 | 56426 | 53961 | ||||||||||||
| 1991 | 23.10 | 21.78 | 56013 | 52811 | ||||||||||||
| 1992 | 21.24 | 20.66 | 52784 | 51335 | ||||||||||||
| 1993 | 19.09 | 19.60 | 48265 | 49536 | ||||||||||||
| The rates from the model can be used to 'backcast' the number of fatalities in past years, and | 1994 | 18.07 | 18.59 | 46537 | 47868 | |||||||||||
| the next graph shows that there is close agreement with the actual number in most years. | 1995 | 17.55 | 17.63 | 46111 | 46333 | |||||||||||
| 1996 | 16.24 | 16.73 | 43521 | 44831 | ||||||||||||
| 1997 | 15.83 | 15.87 | 43413 | 43521 | ||||||||||||
| This degree of agreement provides confidence in the use of this approach to forecast the | ||||||||||||||||
| number of fatalities in future years. To see how to do this, click on the next sheet forecast 1. |
rates
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
forecast 1
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
forecast 2
| 3. Forecasting the fatality rate | actual | modelled | ||||||||||||||
| rate | rate | |||||||||||||||
| There are two steps in forecasting the number of fatalities. Step 1 is to forecast the | ||||||||||||||||
| fatality rate, and the model from the previous sheet provides the natural way to do | ||||||||||||||||
| this. The graph shows how the fatality rate would change up to the year 2010 if the | 1980 | 38.9 | 38.9 | |||||||||||||
| very regular fall since 1970 were to continue. | 1981 | 36.8 | 36.9 | |||||||||||||
| 1982 | 35.0 | 35.0 | ||||||||||||||
| If this were to happen, the rate in 2010 would be 8.0 fatalities per billion veh-km. | 1983 | 34.4 | 33.2 | |||||||||||||
| 1984 | 31.5 | 31.5 | ||||||||||||||
| To see how this information can be used in Step 2 of the forecast, click on the | 1985 | 28.6 | 29.9 | |||||||||||||
| next sheet forecast 2 | 1986 | 28.3 | 28.3 | |||||||||||||
| 1987 | 25.8 | 26.9 | ||||||||||||||
| 1988 | 25.6 | 25.5 | ||||||||||||||
| 1989 | 24.8 | 24.2 | ||||||||||||||
| 1990 | 24.0 | 23.0 | ||||||||||||||
| 1991 | 23.1 | 21.8 | ||||||||||||||
| 1992 | 21.2 | 20.7 | ||||||||||||||
| 1993 | 19.1 | 19.6 | ||||||||||||||
| 1994 | 18.1 | 18.6 | ||||||||||||||
| 1995 | 17.6 | 17.6 | ||||||||||||||
| 1996 | 16.2 | 16.7 | ||||||||||||||
| 1997 | 15.8 | 15.9 | 15.87 | |||||||||||||
| 1998 | 15.1 | 15.06 | ||||||||||||||
| 1999 | 14.3 | 14.28 | ||||||||||||||
| 2000 | 13.5 | 13.25 | ||||||||||||||
| 2001 | 12.9 | 12.28 | ||||||||||||||
| 2002 | 12.2 | 11.39 | ||||||||||||||
| 2003 | 11.6 | 10.56 | ||||||||||||||
| 2004 | 11.0 | 9.80 | ||||||||||||||
| 2005 | 10.4 | 9.09 | ||||||||||||||
| 2006 | 9.9 | 8.43 | ||||||||||||||
| 2007 | 9.4 | 7.81 | ||||||||||||||
| 2008 | 8.9 | 7.25 | ||||||||||||||
| 2009 | 8.4 | 6.72 | ||||||||||||||
| 2010 | 8.0 | 6.23 | 20000 | |||||||||||||
| ETSC forecast | alternative | |||||||||||||||
| 0.949 | 0.927 | |||||||||||||||
| 1.4136137319 |
forecast 2
| 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 |
| 0 | 0 | 0 | 0 |
| 0 | 0 | ||
| 0 | 0 | ||
| 0 | 0 | ||
| 0 | 0 |
| 0 | 0 | 0 |
| 0 | 0 | 0 |
| 0 | 0 | 0 |
| 0 | 0 | 0 |
| 0 | 0 | 0 |
| 0 | 0 | 0 |
| 0 | 0 | 0 |
| 0 | 0 | 0 |
| 0 | 0 | 0 |
| 0 | 0 | 0 |
| 0 | 0 | 0 |
| 0 | 0 | 0 |
| 0 | 0 | 0 |
| 0 | 0 | 0 |
| 0 | 0 | |
| 0 | 0 | |
| 0 | 0 | |
| 0 | 0 |
| 4. Forecasting the number of fatalities | ||||||||||||||||
| 1980 | 1645 | |||||||||||||||
| To convert the forecast rate in 2010 of 8.0 fatalities per billion veh-km into the number of | 1981 | 1661 | ||||||||||||||
| fatalities, a forecast is needed of the traffic growth between 1997 and 2010. The EC's | 1982 | 1705 | ||||||||||||||
| draft 'Communication on the Revision of the Common Transport Policy' of April 2000 | 1983 | 1730 | ||||||||||||||
| forecast that passenger mileage by road would grow by 16% between 1997 and 2010, | 1984 | 1797 | ||||||||||||||
| and that freight transport by road would grow by 19%. This suggests an overall | 1985 | 1840 | ||||||||||||||
| growth in road traffic of about 17%. The graph shows that this would be slower growth | 1986 | 1936 | ||||||||||||||
| than in recent years. | 1987 | 2044 | ||||||||||||||
| 1988 | 2154 | |||||||||||||||
| 1989 | 2255 | |||||||||||||||
| 1990 | 2351 | |||||||||||||||
| 1991 | 2425 | |||||||||||||||
| 1992 | 2485 | |||||||||||||||
| 1993 | 2528 | |||||||||||||||
| 1994 | 2575 | |||||||||||||||
| 1995 | 2627 | |||||||||||||||
| 1996 | 2680 | |||||||||||||||
| 1997 | 2742 | 17% | 1.0121504354 | |||||||||||||
| 1998 | 2775.6 | |||||||||||||||
| 1999 | 2809.4 | |||||||||||||||
| 2000 | 2843.5 | |||||||||||||||
| 2001 | 2878.0 | |||||||||||||||
| 2002 | 2913.0 | |||||||||||||||
| 2003 | 2948.4 | |||||||||||||||
| 2004 | 2984.2 | |||||||||||||||
| 2005 | 3020.5 | |||||||||||||||
| Number of fatalities = | Fatalities | X Traffic = Fatality rate X Traffic | 2006 | 3057.2 | ||||||||||||
| Traffic | 2007 | 3094.3 | ||||||||||||||
| 2008 | 3131.9 | |||||||||||||||
| so the equation needed to forecast the number of fatalities in 2010 is | 2009 | 3170.0 | ||||||||||||||
| 2010 | 3208.5 | |||||||||||||||
| = | 8.0 | X | 3208.5 | |||||||||||||
| = | 25658 | |||||||||||||||
| To achieve this would require continued efforts to improve safety and reduce the fatality rate, | ||||||||||||||||
| roughly equivalent to the efforts made in recent years. | ||||||||||||||||
| The ETSC target of reducing deaths to 25000 in 2010 can be achieved by an extra effort to | ||||||||||||||||
| reduce deaths by a further | 2.6% | |||||||||||||||
| If traffic grows faster than the EC predicts, a greater effort would be required to reach this target. | ||||||||||||||||
| To assess the effect of an alternative assumption about traffic growth, enter your estimate of the | ||||||||||||||||
| % traffic growth by 2010 in the red box. For example, to test the implications of 30% growth, | ||||||||||||||||
| enter 30 and press the Enter key. | ||||||||||||||||
| 25 | traffic growth, 1997 to 2010 | 25% | ||||||||||||||
| traffic volume in 2010 | 3427.9 | |||||||||||||||
| forecast number of fatalities | 27412 | |||||||||||||||
| This needs to be reduced by | 9.6% | |||||||||||||||
| by extra efforts in order to achieve the ETSC target of 25000 in 2010. |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | 0 |
| 0 | |
| 0 | |
| 0 | |
| 0 |
European Commission Proposed Target: 50% Reduction Between 2000 and 2010
*
The case for an EU-wide target has now been accepted by all of the EU institutions and the European Commission has proposed a target to 2010 in the recently published White Paper.
What has actually been proposed by the Commission is a target to reduce deaths to no more than 20,000 by 2010 which represents a 50% reduction compared to a baseline year of 2000.
This goes much further than the previous proposal and anticipates a very very steep decline in the number of deaths.
It implies achieving a safety performance level for the EU as a whole which is much better than that achieved by even the best performing Member States.
This target will require an unprecedented level of activity in the next Action Programme.
Chart4
| 1970 |
| 1975 |
| 1980 |
| 1985 |
| 1990 |
| 1995 |
| 2000 |
| 2005 |
| 2010 |
Sheet1
| Year | Fatalities |
| 1970 | 77989 |
| 1975 | 69510 |
| 1980 | 64063 |
| 1985 | 52695 |
| 1990 | 56426 |
| 1995 | 46111 |
| 2000 | 41137 |
| 2005 | 31000 |
| 2010 | 20000 |
Sheet1
| 0 |
| 0 |
| 0 |
| 0 |
| 0 |
| 0 |
| 0 |
| 0 |
| 0 |
Sheet2
Sheet3
Supplement 2: Writing Policy Memos and Issue Papers
*
POLICY ANALYST
POLICY-MAKING PROCESS
POLICY DOCUMENTS
POLICY COMMUNICATIONS
Interactive Communication
Knowledge Utilization
Policy Analysis
POLICY INFORMATION
Policy Documentation
*
Characteristics of Basic
and Applied Analysis
| Characteristic | Basic Analysis | Applied Analysis |
| ORIGIN OF PROBLEMS | policy scholars | policy stakeholders |
| METHOD OF CHOICE | formal modeling | argument analysis |
| NATURE OF DATA | primary data | secondary data |
| AIM OF ANALYSIS | improve theory | improve practice |
| LOCUS OF INCENTIVES | universities | governments |
*
Criteria for Assessing Policy
Memos and Other Documents
Economy of style
Clarity
Directness
Understandability
Organization
Attention-Getting
Low costs to reader
*
Tasks in Policy Documentation
Synthesizing
Organizing
Translating
Simplifying
Visualizing
Summarizing
*
Elements of Policy Memo
State question(s) the memo will answer
Review prior attempts to solve problem
Diagnose problem scope, severity, causes and identify goals and objectives of problem solution
Compare and evaluate alternatives—benefits and costs and constraints
State conclusions or recommendations
Provide attachments (as appropriate)
*
Elements of Issue Paper
- Letter of Transmittal
- Executive Summary
- Background of Problem
- Scope, Severity, Causes of Problem
- Description of Alternatives--Goals and Objectives
- Analysis of Alternatives—Costs and Benefits
- Conclusions and/or Recommendations
- References/Sources
- Appendices
*
Methods for Creating Information
Needed to Write Policy Memos
BACKGROUND OF PROBLEM
DIAGNOSIS OF PROBLEM
DESCRIPTION OF ALTERNATIVES
ANALYSIS OF ALTERNATIVES
CONCLUSIONS AND RECOMMENDATIONS
Monitoring
Evaluation
Problem Structuring
Forecasting
Recommendation
*
Sequence number
987654321
16
14
12
10
8
6
4
2
0
Probable and
Durable
Probable and
Non-Durable
Improbable
Regression to Mean
Improbable Constant
Change
Improbale Non-Linear
Change
Sequence number
9
8
7
6
5
4
3
2
1
16
14
12
10
8
6
4
2
0
Probable and
Durable
Probable and
Non-Durable
Improbable
Regression to Mean
Improbable Constant
Change
Improbale Non-Linear
Change
Fig. 8.1. Connecticut traffic fatalities, 1955-56
YEAR
1957195619551954
FATALITIES
330
320
310
300
290
280
YEAR
19751974197319721971
FATALITIES
56000
54000
52000
50000
48000
46000
44000
42000
56000
54000
52000
50000
48000
46000
44000
42000
Fig. 8.2. Connecticut traffic fatalities, 1951-59
YEAR
.195919581957195619551954195319521951.
FATALITIES
340
320
300
280
260
240
220
YEAR
2000
1998
1996
1994
1992
1990
1988
1986
1984
1982
1980
1978
1976
1974
1972
1970
1968
1966
FATALITIES
60000
50000
40000
30000
Fig. 8.3. Connecticut and control states traffic fatalities, 1951-59
YEAR
.195919581957195619551954195319521951.
FATALITIES
340
320
300
280
260
240
220
Control States
Connecticut
Transforms: natural log, difference (1)
YEAR
2000
1998
1996
1994
1992
1990
1988
1986
1984
1982
1980
1978
1976
1974
1972
1970
1968
Change in Fatalities
.2
.1
0.0
-.1
-.2
Fatalities
Economic Index
YEAR
76757473727170
28
26
24
22
20
18
16
US
EURCON
EUREXP
Year
1996199019841978197219661960195419481942193619301924191819121906
Fatalities Per Mile Traveled
.4
.2
0.0
-.2
-.4
Billion Miles Driven
Traffic Fatalities
0
10
20
30
40
1980198519901995200020052010
fatalities per billion veh-km
fatality ratemodelETSC forecast
0
10000
20000
30000
40000
50000
60000
70000
80000
197019751980198519901995200020052010
Fatalities per year
